How do I pay for paid tools?
Call `get_invoice(endpoint_id)` → send RLUSD on XRPL mainnet to the returned address with the memo hex → call `verify_payment(invoice_id, tx_hash, agent_wallet)` → receive a 1‑hour access token to use with any paid tool.
S
Squeezeos MCP Server
@timwal78
基本信息
关于 Squeezeos MCP Server
Institutional AI market intelligence for autonomous agents. Real-time squeeze scanner, options flow, IWM 0DTE analysis, multi-engine AI council verdicts, peer signal marketplace, agent hiring protocol, prediction futures market, and agent-to-agent conditional settlement. Pay-per-
Squeezeos MCP Server 配置
使用下面的配置,将此服务器添加到你的 MCP 客户端。
{
"mcpServers": {
"squeezeos": {
"command": "npx",
"args": [
"-y",
"mcp-remote",
"https://lively-fascination-production-41fa.up.railway.app/mcp"
]
}
}
}Squeezeos MCP Server 工具
工具列表暂未提供
当前收录暂时没有可展示的工具列表，不代表服务没有工具。请查阅项目文档了解支持的能力。
点击上方按钮,在临时沙箱中运行此服务器并实时获取工具列表。
概览
What is SqueezeOS?
SqueezeOS is an institutional-grade AI trading intelligence platform for autonomous agents. It offers two live MCP servers with 33 tools total, accessed via pay-per-call using RLUSD on the XRP Ledger (or USDC on Base) through the x402/HTTP‑402 protocol. Designed for AI agents, it requires no API keys, subscriptions, or accounts.
How to use SqueezeOS?
Connect any MCP client (Claude, GPT, etc.) by adding the server URL https://squeezeos-api.onrender.com/mcp with transport streamable-http. For paid endpoints, first call get_invoice to get payment details, send RLUSD on XRPL, then call verify_payment to obtain a 1‑hour access token. A Python SDK is available for automated payment flow. Free discovery endpoints and a live demo (/api/demo/council) let you test responses without paying.
Key features of SqueezeOS
- 33 MCP tools — 15 free, 18 paid via x402
- Pay per call in RLUSD (XRPL) or USDC (Base) — no subscriptions
- Token‑based access: 1‑hour HMAC‑SHA256 token after payment
- Zero simulated data: returns “AWAITING_DATA” if live data unavailable
- Agent Credit Bureau: portable 300–850 score from XRPL spend history
- Multi‑engine oracle: 8 engines (GammaFlow, VPIN, Battle, etc.) aggregated per request
Use cases of SqueezeOS
- AI agents obtain real‑time institutional trading signals with confidence scores and targets
- Agents scan the $1–$50 stock universe for squeeze candidates and options picks
- Peer‑to‑peer signal marketplace: list, read, and stake on signal predictions
- Conditional escrow contracts settled automatically on XRPL
- Regulatory event feed queries (SEC 8‑K, FDA, USPTO) with keyword search
常见问题
What data sources does SqueezeOS use?
Priority order: Tradier (options chain) → Alpaca (OHLCV) → Polygon → Alpha Vantage. All data is live; if unavailable the API returns `status: "AWAITING_DATA"` — never fabricated values.
What runtime or dependencies are required?
You need Python 3, the packages in `requirements.txt`, and optionally `TRADIER_API_KEY` and `PROOF402_TOKEN_SECRET` in a `.env` file for local development. The MCP server runs via Docker/gunicorn on Render.
Is there a free way to try the service?
Yes: `curl https://squeezeos-api.onrender.com/api/demo/council` returns a full council verdict for IWM (5‑minute cache). Fifteen free tools are also available, including `demo_council`, `signal_preview`, `system_status`, and the Agent Credit Bureau score endpoint.
What is the x402 payment flow?
The x402 flow works as follows: `GET /api/{endpoint}` → server responds with HTTP 402 + payment terms → agent pays USDC/RLUSD → retry with `X-PAYMENT` header → server returns 200 with data. For SqueezeOS, the payment asset is RLUSD (issuer: `rMxCKbEDwqr76QuheSUMdEGf4B9xJ8m5De`) on XRPL mainnet.
其他 分类下的更多 MCP 服务器
Yuppsy
Yuppsy MCP: build and publish real estate websites from your AI assistant. Yuppsy lets your AI assistant create live websites for real estate agents and individual property listings. You don't need a designer or any code
Hindsight
vectorize-ioHindsight: Agent Memory That Learns
50.4K
ICSS
chokcoco不止于 CSS
20.7K
FastMCP v2 🚀
jlowin🚀 The fast, Pythonic way to build MCP servers and clients.
26.0K
Sequential Thinking
modelcontextprotocolModel Context Protocol Servers
88.2K
评论